Erlotinib hydrochloride (1), chemically named as N-(3-ethynylphenyl)-6,7-bis-(2-methoxyethoxy)-4-quinazolinamine monohydrochloride, is an inhibitor of oncogenic and proto-oncogenic protein tyrosine kinases, e.g. epidermal growth factor receptor (EGFR). Erlotinib is therefore useful in the treatment of proliferative disorders and is currently marketed for the treatment of lung cancer and pancreatic cancer.

It has been reported that erlotinib hydrochloride can exist in different polymorphic forms. The manufacturing process for many pharmaceuticals is hindered by the fact that the organic compound which is the active ingredient can exist in more than one polymorphic form. It is essential in pharmaceutical development to ensure that the manufacturing process for the preparation of the active ingredient affords a single polymorph with a consistent level of polymorphic purity. If the manufacturing process produces a product with varying degrees of polymorphic purity and/or or where the process does not control polymorphic inter-conversion, it could lead to serious problems in dissolution and/or bioavailability in the finished pharmaceutical composition comprising the active ingredient.
Erlotinib hydrochloride is disclosed in U.S. Pat. No. 5,747,498 and details of the disclosed method for the preparation of erlotinib hydrochloride are described in Scheme 1.

4-Chloro-6,7-bis-(2-methoxyethoxy)quinazoline (2) was reacted with 3-ethynylaniline (3) or its hydrochloride salt using various solvents and pyridine as a base to yield erlotinib hydrochloride (1) which was treated with a biphasic mixture consisting of saturated aqueous NaHCO3, chloroform and methanol, to form erlotinib base (4). The base (4) obtained in the organic phase was purified by flash chromatography to afford purified erlotinib base. The purified base was further treated with hydrochloric acid in the presence of diethyl ether and chloroform to yield erlotinib hydrochloride. This isolation of purified erlotinib base required the use of a lengthy workup process including column chromatography and required the chlorinated solvent, chloroform, which is not particularly suitable for commercial production of pharmaceuticals. Furthermore, the purification by column chromatography is neither economical not feasible at industrial scale. In addition, substantially pure erlotinib could not be obtained.
Two crystalline forms of erlotinib hydrochloride (polymorph A and polymorph B), were characterized by XRPD in patent application, WO 01/34574. Erlotinib hydrochloride can be obtained in form A or in a mixture of polymorph A and B, by refluxing 3-ethynylaniline and 4-chloro-6,7-bis-(2-methoxyethoxy)-quinazoline in a mixture of toluene and acetonitrile. This afforded polymorph A or a mixture of polymorph A and B. It was also disclosed that the formation of polymorph A was favoured by reducing the amounts of acetonitrile with respect to toluene. Furthermore, erlotinib hydrochloride polymorph A can be converted into polymorph B by refluxing the polymorph A with alcohol/water. Consequently, in the disclosed methods, there was always contamination of form A with form B and vice-versa. In addition, the products of the reaction are not chemically pure and difficult to purify thereafter. Consequently, these methods are not suitable for preparation of commercial quantities of pure polymorph A.
A process for the preparation of erlotinib hydrochloride, polymotph E, by condensation reaction of 3-ethynylaniline and 4-chloro-6,7-bis-(2-methoxyethoxy)quinazoline in (α,α,α)-trifluorotoluene and HCl was disclosed in U.S. Patent application 2004/0162300. Polymorph E was characterized by XRPD, IR and melting point. However, (α,α,α)-trifluototoluene is a highly flammable and dangerous solvent for the environment and is not suitable for commercial production.
A process for the preparation of erlotinib hydrochloride, polymorph A by reaction of erlotinib base with aqueous or gaseous HCl was disclosed in US 2009/0131665. In this method, toluene, a mixture of toluene and methanol, TBME, ethyl acetate, 1-butanol or MIBK were used as a solvent. However, when DCM, diethyl ether, isopropyl acetate, was used as a solvent, polymotph B was formed. In practice, it has been found that the disclosed methods are inconsistent and afford polymorphic mixtures. In particular, example 1 of US 2009/131665 was repeated and erlotinib hydrochloride was obtained with only 97% purity. In addition, XRPD analysis showed that the example afforded form B or mixtures of forms A and B. Furthermore, several crystallizations of erlotinib hydrochloride, obtained from repetition of the example, using various solvents and their combinations would not yield a product pure enough to comply with ICH guidelines.
A process for the preparation of a hydrate of erlotinib hydrochloride comprising crystallization of erlotinib hydrochloride using water as solvent, preferably in the absence of organic solvent was disclosed in US 20080167327. This patent also disclosed the process to prepare hemihydrate polymorph form I as well as form II.
A process for the preparation of erlotinib hydrochloride, polymorph M, N and P by reaction of erlotinib base and aqueous or gaseous HCl dissolved in organic solvents was disclosed in WO 2008/102369.
A process for the preparation of erlotinib hydrochloride by condensation reaction of 4-chloro-6,7-bis-(2-methoxyethoxy)-quinazoline and 3-ethynylaniline in isopropyl alcohol as a solvent and pyridine as a base was disclosed in Molecules Journal (Vol. 11, 286, 2006) but no details on the polymorph were disclosed.
A method for the preparation of erlotinib hydrochloride polymorph A comprising passing hydrochloride gas onto solid erlotinib base containing residual amounts of isopropanol was disclosed in WO 2010/040212. However, in practice it was found that the process did not afford chemically or polymorphically pure product. Repetition of example 1 (page 8) of WO 2010/040212 to prepare erlotinib hydrochloride, by reaction of erlotinib base and gaseous HCl in IPA as a solvent, afforded a mixture of polymorph A and polymorph B (as checked by XRPD).
A process for the preparation of acid salts of erlotinib by reaction of 4-chloro-6,7-bis-(2-methoxyethoxy)-quinazoline and 3-ethynylaniline or an acid salt of 3-ethynylaniline under acidic conditions to form the corresponding erlotinib salt was disclosed in US 2010/0094004. In order to complete the reaction, several hours (6 hours) of reflux was required and hence it is not a cost effective process. In addition, in practice it was found that the process did not afford chemically or polymorphically pure product.
A process for the preparation of erlotinib base, polymorph G1, G2 and G3 was disclosed in WO 2009/002538 and WO 2010/05924.

A method for the preparation of erlotinib hydrochloride was disclosed in US 2009/0306377. The method, illustrated in Scheme 2, involves treating 6,7-dimethoxy-4(3H)-quinazolone (5) with hydrobromic acid or pyridine-hydrochloric acid to afford 6,7-dihydroxy-4(3H)-quinazolone (6), which was diacetylated with acetic anhydride to afford diester (7), which was treated with oxalyl chloride/DMF to afford 4-chloro-6,7-diacetoxyquinazoline (8). Compound (8) was condensed with 3-ethynylaniline to afford N-(3-ethynylphenyl)-6,7-dihydroxy-4-quinazolinamine hydrochloride (9), which was converted into the diol N-(3-ethynylphenyl)-6,7-dihydroxy-4-quinazolinamine (10) by treatment with aqueous ammonia/methanol. The diol (10) was treated with 2-iodo-ethylmethyl ether to yield compound (4) which on treatment with HCl afforded erlotinib hydrochloride (1). However, this preparation of erlotinib hydrochloride is a long synthetic route and gives low yields and requires very toxic reagents like pyridine, HBr and controlled reagents like acetic anhydride. Hence, it is not suitable for large scale production.